[Bug 541072] Review Request: certmonger - certificate status monitor and PKI enrollment client

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=541072

--- Comment #3 from Nalin Dahyabhai <nalin at redhat.com> 2010-01-25 13:28:21 EST ---
(In reply to comment #2)
> rpmlint has a few warnings that should be addressed:
> 
> % rpmlint -iv ../RPMS/x86_64/certmonger-0.17-1.fc12.x86_64.rpm 
> certmonger.x86_64: I: checking
> certmonger.x86_64: W: conffile-without-noreplace-flag
> /etc/dbus-1/system.d/certmonger.conf
> A configuration file is stored in your package without the noreplace flag. A
> way to resolve this is to put the following in your SPEC file:
> %config(noreplace) /etc/your_config_file_here
> 
> certmonger.x86_64: W: missing-lsb-keyword Default-Stop in
> /etc/rc.d/init.d/certmonger
> The package contains an init script that does not contain one of the LSB init
> script comment block convention keywords that are recommendable for all init
> scripts.  If there is nothing to add to a keyword's value, include the keyword
> in the script with an empty value.  Note that as of version 3.2, the LSB
> specification does not mandate presence of any keywords.
>
> 1 packages and 0 specfiles checked; 0 errors, 2 warnings.
> 
> The first error is this in the spec file:
> 
> %config /etc/dbus-1/system.d/*
> 
> I looked at a couple of other packages that put files there and they simply
> don't mark them as config files. Not sure who is right here. In any case this
> should use the %{_sysconfdir} macro instead of /etc

I guess if I'm reasonably certain that the names of the interfaces won't
change, the D-Bus configuration files being (noreplace) won't become an issue. 
Switching to %{_sysconfdir} and (noreplace).

> For the second I guess you just need to add Default-Start and Default-Stop to
> the init file.

An empty Default-Start/Default-Stop in the init script doesn't jive with the
packaging guidelines (Packaging:SysVInitScript) on the wiki.  For the moment
I'm assuming that the guidelines trump rpmlint.

> A suggestion not related to the review: rather than %if 0 for pulling in the
> BuildRequires for the make check stage would a variable that could be set from
> rpmbuild make this easier?    

I guess that'd work.  The new logic should make the dependencies and the entire
%check section conditional on whether or not the 'check' macro is defined,
which can be turned on with '--with check' at rpmbuild-time.

.spec file updated.

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.